feat(api): derive Mix waveform datum density from duration (~333 samples/sec, capped/floored) instead of fixed 2048 buckets

DeepDrftContent/Processors/WaveformProfileService.cs

@@ -42,7 +42,10 @@ public class WaveformProfileService
     /// <paramref name="entryKey"/> in <paramref name="vaultName"/> (defaults to
     /// <see cref="VaultConstants.WaveformProfiles"/> when null). Bucket resolution defaults to
     /// <see cref="WaveformProfileOptions.BucketCount"/> (512) when <paramref name="bucketCount"/> is null;
-    /// pass a higher value (e.g., 2048) for the Mix high-res datum. Returns false (and logs) on any
+    /// callers pass an explicit count for higher-resolution data — e.g. the Mix datum derives its count
+    /// from the audio duration (≈333 samples/sec, see <c>MixWaveformResolution</c>) so long mixes are not
+    /// under-sampled. This service is content-agnostic: it captures however many buckets it is told to and
+    /// does not itself decide the count. Returns false (and logs) on any
     /// failure — a missing profile is handled gracefully downstream, so callers on the upload path
     /// log-and-continue rather than failing the upload. Does not throw for expected failure modes.
     /// </summary>